On-site, Non-speculative Failure Diagnosis with CLODS
Rishikesh Devsot, Yi Fan Yu, ChenXing Yang, Ellen Shi, Ding Yuan
2026Year
Abstract
Diagnosing complex production failures is a notoriously difficult task. The bottleneck of diagnosing these failures is often wild goose chases: developers make incorrect hypotheses about the root cause, leading to time and effort spent in verifying these hypotheses, ultimately realizing they were wrong. This is caused by speculations not backed by sufficient evidence because there is insufficient diagnostic data.
Ask about this paper
Ask your agent about it.
Lune has read the top-tier papers around this one, so every answer names the papers it rests on.
Related papers
- Root Cause Analysis of Failures in Microservices through Causal DiscoveryAzam Ikram, Sarthak Chakraborty, Subrata Mitra, Shiv Kumar Saini et al.NeurIPS 2022 · 185 citations
- BugDoc: Algorithms to Debug Computational ProcessesRaoni Lourenço, Juliana Freire, Dennis E. ShashaSIGMOD 2020 · 9 citations
- Live forensics for HPC systems: a case study on distributed storage systemsSaurabh Jha, Shengkun Cui, Subho S. Banerjee, Tianyin Xu et al.SC 2020 · 12 citations
- PerFlow: a domain specific framework for automatic performance analysis of parallel applicationsYuyang Jin, Haojie Wang, Runxin Zhong, Chen Zhang et al.PPoPP 2022 · 10 citations
- Robust Root Cause Diagnosis using In-Distribution InterventionsLokesh Nagalapatti, Ashutosh Srivastava, Sunita Sarawagi, Amit SharmaICLR 2025
